[00:02:22] she may have left. [00:02:35] It's tuesday afternoon, yeah? [00:02:35] catch a train and all that. [00:02:41] I have a 5pm 1:1 with her usually [00:04:20] hrm. [00:05:01] so, i'm cutting assets for jan paul; does anyone know if my transparent images have to be .gif or can be .png? [00:09:00] usually png is fine for transparency these days [00:09:34] not sure what the ie6 compat story is at the moment [00:09:54] but if it's an 8-bit png then it doesn't matter, it works on ie6 anyway [00:10:27] they're black and white, so i'm groovy, then. [00:10:32] thanks! [00:26:54] werdna: you there? [00:40:44] TrevorParscal: (or anyone with deep RL knowledge): could you take a look at http://www.mediawiki.org/wiki/MediaWiki_1.17#Resource_loader and make sure it's right? [00:41:40] i will reword it a bit [00:45:01] robla: what's the technical level of the audience? [00:45:44] TrevorParscal: anyone who might want to install MediaWiki [00:45:52] ok [00:47:02] it's ok to go over some people's heads if there's an important point to make to developers [00:47:14] k [00:49:28] for comparison & emphasis, robla, should the intro mention that it's the first release in n months? [00:50:14] sumanah: perhaps.... TimStarling? [00:51:09] first major release, you mean [00:52:22] there's a conflict between the detailed release notes & the summary -- is the minimum supported PHP version 5.2.3 or 5.1? [00:53:11] sorry -- the top of the release notes says 5.2.3, then == Compatibility == says "MediaWiki 1.17 requires PHP 5.1 (5.2 recommended). PHP 4 is no longer supported." [00:53:18] 5.2.3, Chad changed it yesterday [00:54:07] he tested it with 5.1 and found out it didn't work, so he updated the version requirements [00:54:23] the version requirements near the bottom still say 5.1. [00:55:08] robla: if we're deprecating MySQL 3.23.x support with this release, the summary should include that, right? [00:55:47] yeah, probably [00:55:55] I'll add [00:55:57] fixed in my working copy [00:56:06] PHP 5.1 that is [00:57:49] we're not dropping support for MySQL 3.23 now, that same comment was in the 1.16 release notes [00:58:05] oh, never mind then [00:58:08] *sumanah will revert [00:58:39] done [01:00:16] running 3.23 would take someone pretty dedicated. "I dohn' need nun of yer fancy trans-act-ions or new-fangled gadgets!" [01:01:13] *sumanah laughs [01:01:26] you kids get off my non-ACID lawn [01:03:41] it's probably about time to stop saying that we recommend 4.0 over 5.0, since we've just about finished migrating wikimedia to 5.1 [01:04:06] I was about to ask about that [01:04:56] is someone else making changes to RELEASE-NOTES in 1.17? I don't want to get conflicts [01:05:14] I'm not making changes to that file [01:05:54] "MediaWiki 1.17 ships with a ResourceLoader" [01:06:54] maybe it ships with a resource loader, or a resource loader called ResourceLoader, or maybe a client-side framework called ResourceLoader [01:07:04] probably not "a ResourceLoader" though [01:07:28] yeah, that is pretty awkward [01:07:56] maybe "a client-side framework called ResourceLoader" is best since Trevor's been going around replacing every reference to "resource loader" with "ResourceLoader" [01:08:03] I wouldn't want to offend him [01:08:04] *robla svn ups so he sees the current version [01:11:27] TimStarling: perhaps "MediaWiki 1.17 ships with ResourceLoader: a client-side framework which combines and minifies css and javascript attached to the page." [01:11:52] (I wouldn't worry about offending Trevor though if you just want to say "a resource loader") [01:12:39] I suppose also s/javascript/JavaScript/ and s/css/CSS/ [01:12:46] robla: so, I don't see anything else I have a question about -- anything else you want of me? [01:12:56] sumanah: nope, thanks! [01:13:03] ok, night [01:15:25] robla: http://www.mediawiki.org/wiki/MediaWiki_1.17#Resource_loader [01:16:23] *robla looks [01:18:09] oh no not performant [01:18:29] i couldn't resist, that ones just for you [01:18:30] I know it's your favourite word, trevor [01:18:45] someone back me up here [01:19:53] http://en.wiktionary.org/wiki/performant [01:20:06] yeah I think we've had this discussion before [01:20:15] just because a word is in wiktionary doesn't mean you should use it [01:20:39] heh [01:20:49] #2 google link is "performant isn't a word" [01:20:56] lol [01:21:13] it's just technical jargon - stop trying to force everyone to speak The Queen's English you imperialist! [01:21:21] hah [01:21:35] TrevorParscal: get it added to oxford [01:21:47] as long as no one uses "guesstimate", I'll be fine [01:21:47] <^demon> Is it in the scrabble dictionary? [01:21:53] http://foldoc.org/performant [01:22:22] performifying [01:22:30] sorry, I guess the word didn't become common jargon until after 1995, when that website was created [01:24:22] performography. [01:25:09] maybe we can say it's performant and enterprise-ready [01:27:17] robla: and web scale [01:27:37] I do often hear the word performant [01:27:42] *Ryan_Lane shrugs [01:28:11] not that english matters much anymore. isn't lol in the dictionary now? [01:28:27] saying "performant" is more performant than saying "well performing" [01:29:55] *robla takes a run at editing the description [01:31:51] robla: if it means that much to you, efficient would be fine there too [01:32:09] TrevorParscal: I'm already ahead of you there :) [01:32:41] TrevorParscal: give the new version a read....I shortened it up a bit [01:33:36] "performant" doesn't bug me all that much, but the new prose actually takes advantage of the ambiguity of saying "efficient" [01:34:28] i just fixed a plural mistake and took some more blah text out [01:34:40] and I'm fine with using efficient - let's not offend the delicate grammatical sensibilities of our Australian contingent [01:36:04] <^demon> Or your east-coasters ;-) [01:36:25] ugh, don't get me started on their gibberish! [01:37:05] ^demon: don't harsh my buzz man [01:38:16] i'm outta here [01:38:32] oops, outta isn't a real word! how will I ever live with myself? [01:40:45] TimStarling: I think we're done futzing with the ResourceLoader section now (at least I am) [01:41:21] ok [01:42:02] it doesn't say what it is though, just what it does [01:42:24] oh, good point....I'll copy the text I suggested for the HISTORY file in [01:44:08] done [01:45:35] *robla futzes some more [01:50:17] TimStarling: I took another stab at it. I almost nuked the first two sentences, but decided to put them where they are. [01:50:22] now off to grab some food [01:50:30] bbiab [13:07:10] *guillom summons sumanah. [13:08:02] <^demon> Did you sacrifice a virgin? [13:11:16] Are you volunteering? [13:11:55] <^demon> Wow, awkward. [13:12:02] I did light up a candle, chant and danced the Sumana dance. [13:12:17] dance, even. [13:20:51] *werdna chuckles [13:22:12] werdna, this month I have enough to say about LQT that it has its own paragraph! [13:22:20] *werdna sends to bash [13:22:22] guillom: yay [13:22:27] guillom: can see? [13:22:53] Sure, feel free to improve too: http://www.mediawiki.org/wiki/Wikimedia_engineering_report/2011/April#Discussions_and_Interactions [13:25:53] did so [13:26:48] Oh, and it looks like I had messed up the formatting. [13:26:49] *guillom fixes. [13:27:37] Thanks, werdna. [13:42:25] Dammit, why is wikitech-l becoming so active :( [13:45:09] <^demon> It's time for our bimonthly parser/wysiwyg discussions [13:54:46] http://www.mediawiki.org/wiki/MediaWiki_1.18 is quite funny. [14:35:34] is there a mediawiki function that will get me an article name/link based on a page id? [14:38:33] Title::newFromID() [16:45:29] guillom: wikilove is NOT going away. It's been elevated to an extension; jan paul and kaldari are working on it; i cut graphics for jan paul just yesterday. [16:46:12] WikiLove is part of the glorious 5-year plan ;) [16:46:32] it will be rationed to all WikiCitizens [16:46:59] robla: will you kindly send me a new WMF engineering monthly call invite, to my new @wikimedia.org email? [16:48:32] sumanah: done [16:51:21] Y'see, this was the missing piece. [16:51:27] This is what we've been waiting for. [16:51:52] A quicker way to spam people on their talk pages with templates. [16:53:34] stand-up time [16:53:50] There's no need for stand-up when reality suffices. [16:55:00] hmm, I came here for a stand-up meeting, but I'm the only one here :P [16:55:32] kaldari: then you should probably feel free to sit down [16:55:58] *kaldari sits down [16:56:11] A fundraising gender gap that eats new users. There's going to be an "update from Sue" this month. Are we taking bets on the new looming issue or will the New York Times be determining that? [16:56:58] Myra: you're beginning to sound like Shirley, et al. [16:57:13] Oh dear. [16:57:32] :) [16:57:41] kaldari: hi [16:58:02] howdy [16:59:03] Myra: My bad, I thought you were actually a different person :) [16:59:16] kaldari: is neil online [16:59:19] I'm amused. [16:59:19] can't see him [16:59:41] he's joining [16:59:45] cool [17:00:12] Hi Neil. [17:00:16] hi neil [17:00:48] bonjour [17:00:52] ok, who's first? [17:00:57] flipzagging: first :-) [17:01:03] *kaldari stands-up [17:01:04] kaldari: I'm still working on a way to insert "usability," "editor trends," "whitepaper," and "survey" into my bit. [17:01:17] first? [17:01:45] neilk_: any updates after deploying upload wizard today [17:02:23] kaldari: any updates from your side? [17:02:34] Just a bunch of minor code fixes [17:02:37] we haven't deployed yet, I don't understand [17:02:56] got a bunch of code reviews from neil and roan [17:02:58] neilk: did you not uw to prototype [17:03:06] that's just svn up [17:03:09] not real deploying [17:03:16] neilk: true [17:03:45] well, I got bogged down in reviews & mysterious thumbnailing issues [17:04:06] what's going on with the thumbnailing? [17:04:09] we are going to deploy at 1pm, but chances are it won't include the blacklist-oriented fix. [17:04:14] it doesn't work for videos on commons. [17:04:20] it DOES work on prototype. [17:04:36] it works for video thumbnails on prototype? [17:04:40] I just bit the bullet and installed ffmpeg locally to diagnose it, wasting half the morning [17:04:50] yes, it works for video thumbnails on prototype. [17:04:59] where video = ogg theora. [17:05:03] yup [17:05:30] I don't know why it doesn't work on the cluster. Seems like it could be configuration. [17:05:30] so theoretically it should work for ogg formats on commons too [17:05:43] could mdale help debug that on commons [17:05:50] I don't know, maybe [17:05:55] what happens with audio files? [17:06:16] my first guess isn't mdale but rather someone who can hack into Commons to see what's really going on, ie ops person or Roan, which we've done before. [17:06:20] Haven't tried audio files today. [17:06:25] Hey guys [17:06:32] hi roan [17:06:33] Are you having an UploadWizard standup or something? [17:06:35] howdy roan [17:06:37] yup [17:06:40] yes [17:06:40] OK [17:06:51] That gives me a few mins to finish my dessert before I call into the AFT standup [17:07:04] neilk: could ryan lane help [17:07:21] http://en.wiktionary.org/wiki/stand-up#Noun # Interesting. [17:07:22] any ops person could help [17:07:38] but, I was going to try to simulate the problem locally first [17:07:42] Myra: Huh, I hadn't expected it to be in Wiktionary [17:07:49] since I've never gotten video thumbs working on my desktop anyway. [17:07:56] I wouldn't have expected anyone to use it in the current context. [17:07:57] neilk: ok [17:08:00] So I suppose we're even. [17:08:01] but, the problem here isn't necessarily the problem there. :) [17:08:13] It's being used a synonym for "meeting" here, I guess. [17:08:14] hey, I'm standing up in cyberspace [17:08:15] I probably should ping ops. [17:08:16] neilk: agree [17:08:26] agreed [17:08:37] The ops folks are working on their UploadWizard voodoo doll. [17:08:49] i have talked with ct and rlane so do ping them as needed [17:08:50] Myra: standup specifically means a super-short meeting where you just say what you're doing and if you're blocked. Stand-up since you're not supposed to be seated to increase urgency. [17:09:05] And bloodflow. [17:09:09] yeah [17:09:21] Like Alolita is reminding me not to have tunnel vision about how to fix an ops problem. [17:09:27] It needs more punctuation, I think. Stand-up!!! [17:09:45] ok so [17:09:56] I guess I will work on thumbnailing issues until zero hour [17:10:07] and I'll have to let title issues go :( [17:10:07] neilk: let's touch base later on the thumbnailing issue then [17:10:09] when's zero hour now? [17:10:14] 1pm [17:10:15] 1pm PT [17:11:38] i have to switch to aft now [17:11:50] so let's touch base around 1230 then [17:11:54] alolita: np, thanks for doing IRC [17:11:58] ok [17:12:07] neil, kaldari: tty then! [17:18:28] how does Commons determine that an .ogg file is a video or an audio file anyway? [17:18:45] MIME type field in the DB maybe? [17:18:54] they're both application/ogg right? [17:19:15] Nooo idea [17:19:27] black magic I suppose [17:19:32] `img_media_type` enum('UNKNOWN','BITMAP','DRAWING','AUDIO','VIDEO','MULTIMEDIA','OFFICE','TEXT','EXECUTABLE','ARCHIVE') DEFAULT NULL, [17:20:04] that's an interesting list [17:20:08] drawing? [17:20:28] must be for SVGs [17:20:31] Dude, don't ask, I have no idea where that stuff cmes from [17:29:48] Krinkle: thanks for adding the bit that r uns the sql file when running update.php [17:30:16] awjr: I was about to commit it, but the reality is that Trevor did that a second earlier. [17:30:23] RoanKattouw: Krinkle: awjr: Does this look wrong to you? http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/ArticleFeedback/api/ApiQueryArticleFeedback.php?view=markup#l75 [17:30:29] line 76 in particular [17:30:29] lol well thanks TrevorParscal as well [17:30:35] or, 75 [17:30:56] ok, start at line 74 http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/ArticleFeedback/api/ApiQueryArticleFeedback.php?view=markup#l74 [17:31:14] it should be if ( !count( $userRatings ) ) { [17:31:39] TrevorParscal: Makes sense [17:31:40] indeed [17:32:34] Although the actual API output, TrevorParscal , seems okay [17:32:49] Krinkle: i also modified ArticleFeedback.sql to add an index to the timestamp filed in the articlefeedback table. if that table already exists, would the index get added when update.php is run? [17:32:49] data .query .articlefeedback[0] .status = 'current' [17:33:08] the condition in which this code is run is unique [17:33:14] it's probably being skipped over [17:33:19] awjr: I'm not sure, I think it needs an r87415-ish thing for the index [17:33:31] ok i'll take a look [17:33:32] !r 87415 [17:33:32] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87415 [17:33:51] 'addIndex' presumably, not sure about the exact naming [17:34:29] TrevorParscal: I thinik it's correct actually [17:34:35] that comment should be a line higer imho, [17:34:53] well, I was just doing an edit, rate, edit, rate, edit, rate and then saw "Expired!" [17:34:56] if there no valid ratings, but there is a count(); they are expired. [17:35:01] which seems like something is wrong [17:35:11] hmm [17:35:23] "// No valid ratings exist" refers to the else {} [17:35:30] you are right [17:35:36] hmm maybe i should create that index in a separate file [17:35:42] but I need to figure out why i'm getting expired then [17:35:51] yeahm [17:37:19] awjr: Yes you should [17:38:19] !r 87417 | TrevorParscal [17:38:19] --elephant-- TrevorParscal: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87417 [17:38:53] for running an sql file just containing a 'create index', do i do $updater->addExtenionUpdate( array(applyPatch ... ? [17:38:54] lookin [17:39:18] awjr: There should be an addIndex thing [17:39:24] k [17:39:27] *awjr digs for docs [17:39:27] (instead of applyPatch) [17:39:42] Krinkle: right, if there arent any ratings, there's nothing to be expired [17:44:18] hmm.. now I can't reproduce [17:44:20] odd [17:45:48] im struggling to find documentation about the $updater->addExtensionUpdate stuff, anyone know where i can find more info off the top of their head? [17:47:23] nm found an example in ClickTracking [17:48:56] awjr: the encoding of the JSON blobs seems strange [17:48:59] it's HTML encoded? [17:49:06] " ? [17:49:17] er [17:49:20] i used Format:json [17:49:26] interesting [17:49:27] or JsonFormat:encode [17:49:30] right [17:49:36] FormatJson::encode [17:49:40] or some such [17:49:44] hmmph [17:51:36] locally the json objects appear correct in my db [17:51:48] TrevorParscal ^ [17:51:53] ok [17:52:07] where are you seeing this bad behavior? [17:52:14] also, you should purge before adding in the populate script [17:52:22] i see it in mysqladmin, so it may not be real [17:52:23] brb [17:52:39] purge? the stats table? [17:53:42] i was curious what the best way to handle that was and opted for just storing the data in perpetuity since it could be useful to see the highs/lows over time [17:56:57] hmm [17:57:28] that's probably fine i guess [17:57:35] no reason to purge it [17:57:41] makes sense [17:57:56] we could occasionally dump it and purge it [17:58:07] yeah, and it'd be easy to add in if we decide to later on [17:58:12] for performance, if it got to that point, but it would take a very long time [17:58:36] i got it to show me a row, but now it won't [17:58:44] and I know there's a row to show [17:59:06] howief, ping? [17:59:19] hey [17:59:23] you've got data in article_feedback_stats_highs_lows [17:59:23] ? [17:59:31] howief, hi; got a few minuts? [17:59:31] yes [17:59:36] and it's not showing in the table... [17:59:41] o_O [17:59:49] we're pushing AFT to prototype right now, but what's up? [18:00:14] howief, wanted to check with you when you were planning to post the pre-announcement about AFT 3 [18:00:29] soon :) [18:00:37] as in, today? [18:00:40] no [18:00:43] ok [18:00:48] we're planning on pushing the tool to 100k articles on Monday [18:00:52] (this coming Monday) [18:00:58] so maybe sometime in the next few days [18:01:02] ok [18:01:03] thanks [18:01:31] dario's been working on the research [18:01:51] so as of late yesterday, I think we have enough to start writing [18:02:04] "You are viewing this page on prototype.wikimedia.org, which might be a proxy or phishing site. This site can intercept your password; you are strongly advised to log in from en.wikipedia.org." [18:02:06] eh.. ? [18:02:07] so AFT 3 and UW 1 will both be deployed on Monday. [18:02:16] lol [18:02:20] (tentatively) [18:02:27] i think so [18:02:31] ok [18:02:33] thanks [18:03:09] RoanKattouw: the deployment to 100k articles is on Monday, correct? (Alolita keeps verbally saying Tuesday in meetings) [18:03:12] AGH! [18:03:16] found it [18:03:32] guillom: i was going to try to start drafting something later today [18:03:35] IT'S ON MONDAY :D [18:03:43] "Wikipedia uses cookies to log in users. You have cookies disabled. Please enable them and try again.", what a non-sense. [18:03:45] It was on Tuesday but I moved it [18:03:48] what's wrong with this code? http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/ArticleFeedback/SpecialArticleFeedback.php?view=markup#l208 [18:03:48] i'll ping you once i have something for you to look at [18:03:51] roan: Monday what time [18:03:53] Because I wanta non-jetlagged Trevor [18:03:55] thanks RoanKattouw [18:03:55] 9am-11am [18:04:01] howief, sure. Thanks. [18:04:05] awjr: ?? [18:04:06] 9am Monday - will Trevor be ready [18:04:19] Also, heads up [18:04:24] *TrevorParscal thinks awjr ran out of equals signs [18:04:30] Over in the other channel Ryan's saying we should move ClickTracking to a UDP backend [18:04:35] Which makes infinitely more sense [18:04:41] And makes me feel stupid for not thinking of it [18:04:49] oops [18:04:59] use 3 of em [18:05:09] make sure it knows we really mean business! [18:05:20] rlane had a good suggestion - it may be better to use UDP for click tracking events [18:05:28] are you talking about on line 205, $cache = -1? [18:05:32] I can't log into prototype. Anyone else issues ? [18:05:34] that just needs two [18:05:45] Krinkle: Check remember my login [18:06:17] RoanKattouw: Thx, that worked. [18:06:21] TrevorParscal: the checkbox bucketing doesnt seem to be working for me [18:06:38] in anon state, i opened up all the AFT test pages, some had the checkboxes, other didn't [18:06:57] man i must have the usability initiative stuff set up wrong locally, all of the dependencies and weird paths create all kinds of problems >_< [18:07:11] UsabilityInitiative is history [18:07:25] usability initiative-related exts [18:07:35] Right [18:08:02] RoanKattouw, will you be deploying both uploadwizard & AFT3? [18:08:09] Nooooo [18:08:14] Just UploadWizard this afternoon [18:08:16] (1pm) [18:08:19] AFT is tomorrow [18:08:26] RoanKattouw, no, I mean on Monday (sorry for being vague) [18:08:31] Oh [18:08:32] Yes [18:08:35] ok [18:08:36] thx [18:08:38] Unless Neil says otherwise [18:08:48] AFT is not technically a deployment [18:08:53] But a ramp-up [18:08:54] I am trying not to say otherwise [18:09:04] (Although we may find ourselves deploying some last-minute performance fixes) [18:09:41] RoanKattouw: I like your style of deployment better than Ryan's. He wants me to actually figure out what's going on from first principles, rather than breaking down and doing debugging by print statements :) [18:09:59] guillom: i don't think any communications is necessary for tomorrow's deployment [18:10:08] haha [18:10:08] but definitely necessary for the ramp-up to 100k [18:10:11] (just to clarify) [18:10:13] I'm a shotgun kinda guy [18:10:36] howief, yes, I was just checking if roan was the one deploying both on Monday [18:10:42] RoanKattouw: like he's asking me where uploaded files go on the cluster and I'm like... .uh.... [18:10:44] yup [18:10:54] neilk_: Dude, get shell access and use eval.php [18:11:35] Just suggested that in the other channel [18:11:37] neilk_: ;) [18:11:53] I don't like doing print statements in code [18:11:58] call me crazy [18:11:59] Anyone seen sumanah recently? [18:12:03] I prefer to avoid it if I can [18:12:07] Ryan_Lane: you're not crazy [18:12:09] But I'm not afraid to use them if I have to [18:12:20] I use them when I'm developing all the time :) [18:12:35] guillom, how recently? She was about yesterday [18:12:36] jorm: are you wfh today [18:12:45] Reedy, ok, I must have missed her. [18:12:54] yes. i wfh every wednesday. [18:13:58] kaldari, got a few minutes to add what you did this month on WikiLove to http://www.mediawiki.org/wiki/Wikimedia_engineering_report/2011/April#Discussions_and_Interactions please? [18:14:06] gents, does this look sane: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87424 [18:14:16] TrevoarParscal, Krinkle, RoanKattouw: ^ [18:14:34] guillom: hi [18:14:37] awjr: Just saw it, looks sane. [18:14:40] hi alolita [18:14:48] guillom: I also did some work on Parser Functions, should I add that somewhere? [18:14:56] kaldari, how much work? :) [18:15:21] added multi-language support for #time and fixed some bugs with it as well [18:15:33] kaldari, add it to "other projects" in the feature section [18:15:35] ok, i cant quickly get update.php to run locally without barfing due to usability initiative-related exts wonkiness, so i'm not positive about the addIndex stuff I added in the hooks file [18:16:18] kaldari, while you're at it, if you want to add to http://www.mediawiki.org/wiki/Wikimedia_engineering_report/2011/April#Special_projects , be my guest; tomasz has been a bit busy. [18:18:17] guillom: are you in sync with howie on aft communications - i saw your discussion earlier [18:18:40] alolita, mostly; he'll ping me when he has news. [18:18:58] TrevorParscal: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87425 [18:19:06] did you get things to work correctly locally? [18:19:07] guillom: ok cool ; as you know we're deploying the dashboard tomorrow [18:19:22] awjr: yes [18:19:25] +1 [18:19:31] guillom: and ramp-up to 100k articles on monday [18:19:38] alolita, about that dashboard: is there any documentation posted anywhere? I wanted to link to it in the report but couldn't find it. [18:19:42] ok, should we try putting this on prototype? [18:20:13] guillom: there is a documentation - some requirements drafted by howie which i can send to you [18:20:27] alolita, I meant posted publicly :) [18:20:29] awjr: see http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87426 [18:20:35] rounded numbers are good [18:20:46] guillom: nothing up-to-date on mw.org so far [18:20:50] ok [18:21:18] alolita, I hope to work with you in Berlin, and possibly later in SF, to post a lot of missing documentation to mw.o [18:21:39] guillom: great - let's do that in berlin [18:21:43] TrevorParscal: looks good [18:21:46] guillom: when do you get there [18:22:01] alolita, same time as you guys iirc, except I won't be jetlagged =) [18:22:11] just pissed off with public transport ;) [18:22:23] guillom: are you not there in person - its closeby for you [18:22:25] Reedy, nah, I'm taking the plane this time. [18:22:35] Planes are still public transport :P [18:22:47] reedy: hi! [18:22:53] alolita, oh I'll be there in person, but I don't have 9 timezones to deal with :) [18:22:54] hey alolita [18:23:00] awjr: i'm getting "Object of class ResultWrapper could not be converted to int in /var/www/phase3/extensions/ArticleFeedback/SpecialArticleFeedback.php on line 224" [18:23:07] guillom: cool [18:23:14] Reedy, right. They're much less awful than 20 hours of train tho :) [18:23:18] heh [18:23:28] guillom: so on the upload wizard - we're planning to switch over on monday too [18:23:41] alolita, yep, looking at the announcement right now [18:23:47] guillom: moka is coordinating with us [18:23:50] TrevorParscal, awjr: In that case someone's trying to cast a DB result to an integer [18:23:59] alolita, oh? I wish I knew that kind of things. [18:24:14] its the same line he just fixed - which has the assignment instead of comparison [18:24:17] oh? [18:24:20] guillom: did erik give you an update [18:24:26] he's checking a db result being -1 [18:24:37] alolita, he shared the doc with me, that's all. [18:24:54] ... let me rephrase he's checking if a db result is -1 [18:24:56] TrevorParscal: i copied your code from FundraiserStatistics_body.php :p [18:25:25] guillom: so the upload wizard post will be on blog.wmf - so moka has reviewed it [18:25:26] it's supposed to be checking whether or not there's a valid object in the cache, right? [18:25:31] yeah [18:25:33] what would be the correct way to do that? [18:25:37] RoanKattouw: what's the proper way? [18:25:57] Not compare an object to -1? It's not clear what you even want to do [18:25:58] guillom: anyways ping me if you need any details [18:26:16] alolita, no problem -- I just thought that was my job -- but if you find other people to do it, all the best, as far as I'm concerned :) [18:26:22] awjr: i think I was stashing an integer [18:26:25] no tsure [18:26:31] no, database result [18:26:32] but meh [18:26:33] i was probably making a mistake [18:26:44] wgMemc->get() should just return false if there's nothing in the cache? [18:26:48] we can probably just drop the 2nd comparison alltogether [18:26:50] yeah [18:26:52] im on it [18:27:03] guillom: erik is coordinating that effort [18:27:17] alolita, do note, I'm not complaining =) [18:27:18] awjr, use the source, luke!# [18:27:23] * Get an item with the given key. Returns false if it does not exist. [18:27:27] thanks Reedy :p [18:27:31] Or null, in some cases, I think [18:27:31] Fuck yeah, star wars joke on the 4th May [18:27:33] *Reedy grins [18:27:35] Unless that's fixed now [18:29:33] :D [18:30:32] awjr: you should use: "if ( $cache instanceof ResultsWrapper ) {" [18:30:54] otherwise garbage could end up being assumed to be correct just because it's not false [18:30:57] that is much better and makes me happy [18:31:02] thanks TrevorParscal [18:31:07] n/p [18:32:53] TrevorParscal: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87430 [18:34:42] I made some more visual changes [18:34:45] update at will [18:34:52] k [18:34:53] this looks good to me [18:35:07] in /home/wikipedia/common/php/extensions/ArticleFeedback yes? [18:35:12] Yes [18:35:23] oh - is that supposed to be coming from the wmf branch? [18:35:26] rather than trunk? [18:35:49] ? [18:35:54] Are we talking cluster or prototype? [18:36:02] prototype [18:36:20] Then it's /srv/org/wikimedia/wikis/rc-en/extensions/ArticleFeedback IIRC [18:36:21] awjrichards@prototype:/home/wikipedia/common/php/extensions/ArticleFeedback$ svn info [18:36:21] Path: . [18:36:21] URL: svn+ssh://svn.wikimedia.org/svnroot/mediawiki/branches/wmf/1.17wmf1/extensions/ArticleFeedback [18:36:27] oh [18:37:03] TrevorParscal: It's probably better to just use === -1 , doesn't that work? [18:37:35] Or well, depends on what you ultimately want to check for I guess [18:37:39] instanceof is best I think [18:37:52] because we want to know, is this a ResultsWrapper, or not [18:38:02] i agree with TrevorParscal [18:38:17] awjr: but to Roan's point, I suggested you use 3 equals signs [18:38:28] OK, just saying, what happens when the name of that class change [18:38:31] but i wanted to check if it wasn't -1, not if it was [18:38:32] which would have not tried to do type conversion [18:38:36] Besides, query errors are fatal anyways [18:38:42] awjr: !== [18:38:57] yes :) [18:39:02] RoanKattouw: if the class name changes, grep will most likely be used on the extensions directory [18:39:05] So unless you've specifically disabled error reporting, a failed query will throw a DBErrorException [18:39:06] !== != === [18:39:06] --elephant-- I don't know anything about "==". [18:39:08] TrevorParscal: True [18:39:13] heh [18:39:35] So generally you can just assume select() returns something iterable [18:39:47] And won't return on error [18:40:06] ok so i svn up'd article feedback on prototype, now we need to run update.php to get the sql changes and adjust local settings to turn on dashboard, yes/ [18:40:07] ? [18:40:31] Yes [18:40:37] But update.php is kinda special on prototype [18:40:43] tell me more [18:40:52] Run it with rc_enwiki as the first argument (I think) [18:41:02] (We have a multi-wiki, same-source setup there) [18:41:07] --wiki=rc_enwiki? [18:41:09] No [18:41:17] That's what you use when it's properly set up, such as on the cluster [18:41:20] update.php rc_enwiki [18:41:20] Instead, my setup is a hack [18:41:23] Yes [18:41:27] Use that [18:41:45] update.php rc-en [18:41:51] that did it [18:42:29] ok, I need to get some food in me [18:42:35] neilk_: Alright, UW diff looks good to me [18:42:37] am I needed in the next 60 minutes? [18:42:42] Now moving on to awjr's code [18:42:49] RoanKattouw: thanks [18:42:53] TrevorParscal: If you wanted to talk performance with me, that'll have to happen in the next 60 mins [18:43:02] RoanKattouw: i did a little review there, but left most of the caching and database stuff for you [18:43:10] Or maybe during a few stolen moments tomorrow [18:43:18] Otherwise we won't be able to talk about it until you get here [18:43:22] ok, well I will probably be back before then [18:43:29] Alright [18:43:30] I will catch you in the latter bits of the 60 minuts [18:43:32] cya soon [18:43:35] Alright [18:43:48] ha, Gmail now thinks mail from Jay Ashworth is spam [18:43:49] hilarious [18:44:00] RoanKattouw: where is the local settings config stuff on protoype? [18:45:01] wikis/rccommon/PrototypeSettings.php [18:45:03] wikis/rccommon/ [18:45:05] cool [18:45:39] awjr: what he said :) [18:46:44] http://prototype.wikimedia.org/release-en/Special:ArticleFeedback [18:46:52] TrevorParscal: still around? ^ [18:47:29] can people rate those test pages so that we exceed the 10-rating limit? [18:47:52] oh i suppose those are because messages aren't populated [18:48:52] howief: on it [18:49:02] thanks [18:49:04] awjr: ??? [18:49:29] TrevorParscal: are there not messages for the column headers in the dashboard table? [18:49:29] i can haz i18n? [18:49:43] ok [18:50:03] so, basically the problem is that the contents of the article_feedback_ratings table is wrong [18:50:09] on the cluster we have fixed it [18:50:17] and most of us have it correct on our localhost [18:50:30] but we never made a migration when we changed the names of the columns [18:50:34] and then we changed the keys [18:50:34] oh i see. [18:50:36] so [18:50:47] just look at the contents of your localhost table (4 rows) [18:50:53] and make sure the prototype has those [18:51:23] we should probably add a migration for this, but this is likely the last computer on earth with the wrong data [18:51:59] *TrevorParscal leaves again [18:53:48] yay that worked, thanks TrevorParscal [18:55:00] Hi alolita [18:55:11] mmerz: hi [18:55:25] mmerz: should we wait for denis [18:55:46] No, it will be only us two today [18:56:36] So lets roll [18:56:57] ok [18:57:09] ok i did 2x ratings for each of those pages, howief [18:57:21] great thanks [18:57:27] looking quickly over the article_feedback table we should have enough to go [18:57:36] im going to go ahead and try running the maintenance script [18:57:39] great! [18:59:12] RoanKattouw: is memcache not on prototype? [18:59:27] memcached should be [18:59:36] likely a different port though [18:59:48] different port than the cluster, that is [18:59:57] eval is great [19:00:05] I had no idea MediaWiki had a REPL [19:00:12] It should be configured properly though [19:00:18] neilk_: REPL? [19:00:24] awjr: Run eval.php and play with $wgMemc->get() and set() [19:00:37] oo [19:00:50] eval.php is teh awesome [19:00:54] Totally [19:01:01] RoanKattouw: http://en.wikipedia.org/wiki/Read-eval-print_loop [19:01:10] It'd be nice if home/end etc would work in it [19:01:25] wfm [19:01:31] Except for delete [19:01:41] That just inserts ~ characters [19:01:46] ^[[1~^[[4~> ~ [19:01:49] Yeah, home/end do for me [19:02:01] except, do I need to be in another directory to use it properly? [19:02:12] No, why? [19:02:22] What's not working properly for you now? [19:02:44] RoanKattouw: memcached does not appear to be properly configured in mediawiki on protoype [19:02:48] I tried "return $wgNamespaces" and saw nothing [19:02:54] neilk_: var_dump [19:02:56] var_dump( $blah) [19:03:02] return doesn't work [19:03:04] oh, the instructions say... [19:03:06] nm [19:03:08] You have to explicitly var_dump or print/echo [19:03:19] It's not strictly a REPL [19:03:22] More like a REL [19:03:28] I take it back :) [19:03:57] It doesn't print the return value of the last statement like Firebug does, you have to tell it to [19:04:22] hmm fatal error on special page: http://prototype.wikimedia.org/release-en/Special:ArticleFeedback [19:04:23] RoanKattouw: then the docs are lying [19:04:29] sofixthem [19:04:57] Another cool thing is you don't have to global $wgFoo; anything [19:05:04] Your code runs in the global scope [19:05:49] * To get decent line editing behavior, you should compile PHP with support [19:05:50] * for GNU readline (pass --with-readline to configure). [19:05:56] *Reedy looks [19:06:43] TrevorParscal: sorry to drag you back if you're paying attn ^ [19:07:00] Fatal error: Call to a member function isKnown() on a non-object in /srv/org/wikimedia/prototype/wikis/rc/includes/Linker.php on line 180 [19:07:00] Aha [19:07:01] #341868 - php5-cli: missing readline extension - Debian Bug report ... [19:07:01] From: Ashar Voultoiz [19:07:25] coming from line 103 of SpecialArticleFeedback.php [19:13:13] Linker::link( $target ) - $target should be a Title object, yes? [19:13:50] awjr: i have to run to something [19:14:10] Yeah [19:14:16] i'll check back in with you in an hour or so [19:14:20] $pageTitle = Title::newFromId( $page['page'] ); [19:14:26] no sweat - i should've been done with this at 12 but i pushed back my civicrm deployment til 1pm, but at 1pm i won't be able to work on this potentially for the rest of the day [19:14:30] I'd guess that needs evaluating [19:14:33] howief: k [19:15:08] Reedy: we're passing a Title object as $target, but gett Fatal error: Call to a member function isKnown() on a non-object in /srv/org/wikimedia/prototype/wikis/rc/includes/Linker.php on line 180 [19:15:15] any thoughts? [19:15:33] i'll keep digging but curious if im missing anything obvious [19:15:42] wfProfileIn( __METHOD__ . '-checkPageExistence' ); is 180 for me.. [19:15:46] Are you sure your Title object is really a Title object and not null? [19:15:47] wait, that's trunk [19:15:50] bleh [19:15:53] RoanKattouw: yse [19:15:56] *yes [19:16:01] yeah, this is on prototype [19:16:14] It's $target->isKnown() where $target is the passed in title [19:16:22] Bets are on that it's not a valid title for whatever reason [19:16:39] >_< [19:16:55] http://pastie.org/1865336 [19:17:00] var_dump of the title object ^ [19:17:01] heh [19:17:01] isKnown [19:17:04] bah [19:17:15] Though, one would expect it would've barfed before then due to $row['page'] = Linker::link( $pageTitle, $pageTitle->getPrefixedText() ); [19:17:23] yah [19:17:36] Oh, watch yourself there [19:17:42] Linker::link() is non-static in 1.17wmf1 [19:17:53] oh [19:18:00] i see [19:18:11] Also double-check it's declaration in Linker.php for any param incompat issues [19:18:25] params look ok for LInker->link() [19:18:33] (You have to do $sk->link() where $sk = $wgUser->getSkin() typically) [19:18:43] muh this is uncharted territory for me - so we need to get a link object [19:18:49] oh there it is thanks Roan [19:18:58] A Linker, yes [19:18:58] werdna said this like four hours ago. ;-) [19:19:05] Ah indeed he did [19:19:09] Myra: indeed [19:19:12] This has been refactored to be nicer in trunk [19:19:17] in fact 3 people told me three different things :) [19:19:21] Oh, doh [19:19:24] MediaWiki sucks. [19:19:28] awjr: Are you sure memc is what's broken instead of + if ( !in_array( $page_id, array_keys( $highest_and_lowest_page_ids ))) { [19:19:35] Ahm, paste fail [19:19:38] RoanKattouw: no [19:19:38] + if ( $cache != false && $cache = -1 ) { [19:19:53] wait, no, i am i tried futzing with $wgMemc in eval.php [19:19:57] Note the weak comparison in != false that matches almost everything, followed by an assignment (!) to -1 [19:19:59] was getting empty bag o stuff objects [19:20:05] RoanKattouw: we changed that [19:20:11] Oh [19:20:44] Yeah, you fixed the assignemnt [19:20:47] in r87424 [19:20:56] But not the weak comparisons? [19:21:49] RoanKattouw: long story but see r87433 [19:21:51] no [19:21:54] r87430 [19:22:04] https://secure.wikimedia.org/wikipedia/mediawiki/w/index.php?title=Special:Code/MediaWiki/87430&path=%2Ftrunk%2Fextensions%2FArticleFeedback [19:22:19] Wait, what [19:22:24] You're putting a RESULTWRAPPER in the cache? [19:22:31] is that a nono? [19:22:35] You could at least call iterator_to_array() on it before you do that [19:22:46] Well, it's an unnecessary amount of data [19:22:52] indeed [19:22:54] And TBH I'm not entirely sure it won't blow up in your face [19:22:58] k [19:23:07] iterator_to_array is good call [19:23:22] and then we cah check if ( !is_array( $cache [19:23:25] s/cah/can [19:25:07] Exactly [19:25:26] In fact [19:25:33] Why not store $high_lows in the cache [19:25:44] Avoid doing the processing (including JSON decoding) on every hit [19:26:21] laziness [19:27:10] huh. googledocs keeps track of files you've viewed but haven't been explicitly shared with you? [19:27:15] i just found this in my list: https://spreadsheets.google.com/ccc?key=0AnEbcGiqzNb9dC1remNWelRjdVVIMFdfZndEbkZteGc&hl=en#gid=0 [19:27:51] (it's a neat spreadsheet showing chromatic wavelengths for sound pitches) [19:28:08] RoanKattouw: for now im just going to do iterator_to_array and store that - later i'll fix it to just store highs_lows (it's a good idea, i just dont have the time to implement it smartly right now) [19:28:20] alolita: ping [19:28:30] hi neil [19:28:32] OK [19:28:38] You have more things to do anyway [19:28:46] There's a stream of CR comments in your inbox and I'm not done yet :) [19:28:51] fun [19:28:52] thanks [19:28:54] :) [19:29:02] are these show-stoppers for deployment? [19:29:20] Some of them are [19:29:21] k [19:29:35] i'm almost certainly not going to have time to fix them today [19:29:35] I really really hate code that even /looks/ like it might be insecure DB-wise [19:29:39] neilk: how are things [19:29:53] neilk: ready to roll at 1p :-) [19:29:56] alolita: so, we didn't make a lot of progress this morning. I got access to the live systems to figure out the thumbnail problem [19:30:18] neilk: ok [19:30:25] alolita: so, mostly, it's the changes we had as of 9:00am this morning (which I think Roan has okayed, and almost all of them are "ok" status anyway) [19:30:35] I've looked over everything about an hour ago [19:30:39] That all looked good [19:30:51] alolita: so, once again exciting new features delayed to monday :( [19:31:00] or rather exciting new error recovery delayed [19:31:04] neilk: why not tomorrow [19:31:11] and then monday again [19:31:14] do we have a push scheduled for tomorrow?! [19:31:39] neilk: well we pushed off from this monday to today - so could push to tomorrow [19:31:43] for this week [19:32:27] neilk: i would do this twice rather than switch to primary and push everything else at the same time [19:32:38] switch to primary? [19:32:43] I don't know what that means. [19:32:47] oh I see what you mean [19:32:54] neilk: glad you got it [19:32:57] where UploadWizard > Special:Upload [19:33:02] yup [19:33:27] I'm reluctant to say yes because I don't like committing to having prod ready features in less than 24 hours. [19:33:38] I don't think this is a helpful thing to ask for [19:34:00] no, no, no, no [19:34:02] Not tomorrow [19:34:10] I am not deploying anything tomorrow [19:34:16] roan: ok :-) [19:34:51] neilk: i would still recommend that we push out the minor stuff we have so far [19:34:57] alolita: is there something I've done in the past few weeks where I gave you the impression we had a stable codebase [19:34:57] today [19:35:12] Ahm, well, AFT is tomorrow [19:35:13] alolita: I am trying to meet the deadlines [19:35:17] So I'm deploying that but not UW [19:35:19] RoanKattouw: i don't think so [19:35:20] alolita: there are limits to what I can do [19:35:22] re AFT tomorrow [19:35:29] Oh, OK [19:35:30] neilk: no [19:35:35] Then when? [19:35:36] i think we're going to have to put it off [19:35:40] neilk: understood [19:35:48] i dunno, perhaps we can munge the dashboard deploy with the ramp up [19:35:58] which is ugly but i dunno how else we can do it before berlin [19:36:07] I guess that could be done [19:36:20] neilk: if we deploy on monday as scheduled will that give you adequate time [19:36:21] but im almost certainly not going to get fixes done until sometime tomorrow [19:36:29] and then we still need to test on prototype [19:36:40] alolita: anyway I'm fine doing a regular push, we have a few new things including new header etc. Everything else that we wanted only kicks in during error conditions anyway. [19:37:16] neilk: understood [19:37:37] neilk: so you are recommending that we only push on monday now [19:37:41] no [19:37:49] I'm recommending we push as we planned, now [19:37:54] and then again on Monday [19:38:11] neilk: and we push what we have so far now? [19:38:15] yes [19:38:21] neilk: works for me [19:38:30] ok [19:38:35] neilk: if it works for you [19:38:38] RoanKattouw: I'll be back in < 15 minutes for the push [19:38:44] OK [19:38:47] alolita: sure [19:38:48] neilk: thanks! [19:39:20] alolita: we're going to need to delay the AFT push to deployment [19:39:27] er, push to production rather [19:39:37] awjr: to monday :-) [19:39:43] yes :) [19:39:54] we need time to fix issues [19:39:56] awjr: Please update the WMF engineering calendar [19:39:59] awjr: i agree [19:40:04] and then finish testing on prototype [19:40:33] RoanKattouw, alolita: any reservations about munging dashboard deploy with rampup? [19:40:40] Not from my side [19:40:46] We have liberally allocated time for it [19:40:57] i agree [19:41:16] neilk_: I hope you won't be too offended if the AFT deployment immediately before next Monday's UW deployment runs overtime and we have to push it back to be during lunchtime [19:41:25] awjr: well we need to get howie to test and we need to have time to code fix before any deployment [19:41:35] roan : is n't this fun :-) [19:41:54] I know :) [19:41:59] alolita: yep, assuming my civicrm upgrade goes well, i should have time late this afternoon and tomorrow to work on fixes [19:42:03] roan: i like your use of "liberally allocated time" :D [19:42:40] I meant something else, but that was also true [19:42:41] awjr: no problem; as soon as i get your note i can update howie [19:42:52] I misspoke and meant we had allocated more time than we were probably going to need [19:42:55] "royally" was the word [19:43:02] ha [19:43:08] roan: a freudian slip :D [19:43:13] Exactly [19:46:32] awjr: waiting for your note [19:46:39] alolita: writing it [19:46:49] awjr, so, are you done with the deployment attempt? [19:46:53] awjr: thanks :-) [19:47:01] guillom: we postponed it to monday [19:47:08] guillom: yes, although it was just deployment to prototype [19:47:35] guillom: we plan to push to production on monday [19:48:01] awjr, would you have a few minutes to add some information about what happened this month in fundraising / mobile / offline (to the extend of what you know) ? Tomasz didn't have time to do it and we're going to review the report in 5 minutes: http://www.mediawiki.org/wiki/Wikimedia_engineering_report/2011/April#Special_projects [19:48:16] alolita, ok, thx [19:48:40] guillom: one sec i gotta wrap up this email [19:48:50] sure [19:51:18] awjr: For your reference: https://secure.wikimedia.org/wikipedia/mediawiki/w/index.php?title=Special:Code/MediaWiki/status/fixme&author=awjrichards [19:51:31] long url is long [19:51:36] thanks Roan, you rule [19:55:25] guillom: just added a bit for offline [19:55:34] hopefully the civicrm upgade will be completed today :p [19:56:14] guillom you might ping kaldari about the Kiwix UX study, he'll probably know the current status [19:56:22] unfortunately i have no idea what's going on in the world of OpenZim [19:56:25] awjr, thanks! [19:56:53] lemme see if ic an add something for mobile [19:58:49] neilk_: Alright, deployment time [19:59:50] Just verified there are no new commits I haven't seen, now merging [20:12:27] neilk_: Alright. live on test [20:12:35] ok [20:12:35] kaldari, do you think you can help with the kiwix stuff at http://www.mediawiki.org/wiki/Wikimedia_engineering_report/2011/April#Offline too ? [20:12:46] RoanKattouw: oh wait, I forgot to mention, there is a config change needed [20:12:56] Oh OK [20:13:03] not a big deal [20:13:34] if $wgUploadWizardConfig, add the key-val pair 'altUploadForm' => 'Special:Upload' [20:13:51] and BTW that will be different when we deploy to Commons, it will be 'Commons:Upload' [20:14:10] Ugh [20:14:14] OK [20:14:43] More if ( $wgDBname == 'foo' ) switches in CommonSettings then [20:14:57] I guess [20:15:13] maybe I can write this sort of thing myself now that I have fenari access [20:15:16] in the future [20:15:28] Yeah, you should learn that too [20:16:08] What you /can/ do now [20:16:17] Is check out the changes I made with svn di wmf-config [20:16:18] the logic there would be 'altUploadForm' => ( $wgDBname == 'commonswiki' ? 'Commons:Upload' : 'Special:Upload' ) [20:16:22] Yeah [20:16:25] assuming you don't hate ?: [20:16:29] Right now it's != 'testwiki' [20:16:33] Using an if statement [20:16:34] oh [20:16:36] k [20:17:08] I don't have a deployment branch ready [20:17:20] On fenari I mean [20:17:29] Our config is not in public SVN [20:17:55] so where is wmf-config ? [20:18:11] In /h/w/c/php [20:18:19] ( /home/wikipedia/common/php ) [20:19:30] I think it's a bit backwards [20:19:38] normally it's Special:Upload [20:19:43] only on Commons is it Commons:Upload [20:19:50] I guess it is [20:19:56] I believe the dbname for commons is 'commonswiki' ? [20:19:58] I'll switch it around [20:20:00] Yes [20:20:05] I can edit it right now... [20:20:05] Or you go switch it around then [20:20:24] ... wait. Special:Upload is Commons:Upload on commons? That's. . . [20:20:25] well. [20:20:47] Dude [20:20:56] That's the tip of a very large iceberg [20:21:10] I mentioned to Erik the other day that commons should probably be forked. [20:21:16] Did you know they have implemented different upload forms with Special:Upload?uselang=some-made-up-pseudo-language-code ? [20:21:21] And he didn't take that with the revulsion I expected. [20:21:34] jorm: that's interesting [20:21:39] <^demon> People said that 7 years ago when commons was launched. [20:21:46] <^demon> There's actually a long-sinced WONTFIX'd bug for it. [20:21:58] it's probably the right thing to do, tbh. [20:22:41] RoanKattouw: ok try this out [20:22:44] also: it's seriously hot here in oaktown. [20:22:46] jorm: what do you exactly mean when you say forked? removed from the wmf website properties? [20:22:50] 80 degrees. [20:22:55] forked from mediawiki. [20:22:58] jorm: oaktown :D [20:22:59] <^demon> https://bugzilla.wikimedia.org/3712 [20:23:00] RoanKattouw: all the 'subhead' things are now optional, nothing happens if you don't have them configured. [20:23:16] commons isn't really a "wiki". [20:23:23] RoanKattouw: and feedbackPage and altUploadForm are now subhead thingies [20:23:24] but then, wiktionary could probably use a fork, too. [20:23:34] neilk_: OK, I've just been doing what you told me, so feel free and edit with prejudice [20:23:36] oh crap, one more thing to config, I think [20:23:49] jorm, I think someone tried that already (wikitonaryZ/omegawiki/whatever its called) [20:23:50] oh no nm [20:24:02] RoanKattouw: nm, it's already in the default module config so we're good [20:24:10] listen. i worship no god but Apollo and he's probably pissed at me right now for doing my work in the shade. [20:24:17] <^demon> jorm: Well, MediaWiki is only good for writing encyclopedias. But over time we've shoehorned it into doing lots of other things too :) [20:24:25] neilk_: You done editing config now? [20:25:02] RoanKattouw: yes, it seems to work on test, too [20:25:14] RoanKattouw: that is, this new stuff which configures the subhead. [20:25:17] Let me do the rest now [20:25:17] OK [20:25:19] <^demon> jorm: "MediaWiki is useful for everything, it can simulate nuclear weapons blasts or feed your cat" --TS [20:25:41] hrm. i don't think i'm going to get around to making graphical comps for the MoodBar today. but the document will be done. [20:26:00] this "little" project keeps getting "bigger" [20:26:29] RobH: I played a lot of Portal 2 last night. [20:26:32] are you playing it? [20:26:41] I should be [20:26:47] My brother and sister are [20:26:50] I've finished it [20:26:51] But, work, school, travel [20:26:57] jorm: Yep! completed both single and coop [20:27:03] but will totally play coop again =] [20:27:06] jorm, I was up to 4am playing co-op on Monday night/tuesday morning [20:27:19] i've heard co-op is a non-repeatable experience. [20:27:28] It is [20:27:29] Achievements [20:27:33] meh, its portal, i would play it more than once. [20:27:37] and i expect i'll have to play it with my girlfriend. [20:27:41] maybe not over and over [20:27:43] When you've gone through it, you've got access to all the types on demand [20:27:43] ahh [20:27:48] <^demon> I'm about halfway through co-op right now. [20:28:02] i think i'm on test chamber 19 in single player. [20:28:04] maybe 20? [20:28:13] There is no cake :( [20:28:16] good. [20:28:20] then i win 100 bucks. [20:28:21] RoanKattouw: everything looks ok except one bug [20:28:24] i have a bet about that. [20:28:36] i wish there had been a nod to the cake [20:28:36] RoanKattouw: I can fix it fast, I think, but I thought I *did* fix this. Just a missing message in Hooks [20:28:37] but no cake [20:28:45] the nod was a scribble on a wall, i guess. [20:28:48] OK go ahead [20:28:57] i must have missed it [20:29:02] i got most the achives on the first playthrough [20:29:07] i like that it's showing that there were, like, HUNDREDS if not THOUSANDS of other "test subjects" [20:29:10] merely by my insistence on poking aroudn everyplace [20:29:23] heh [20:29:28] I need to do another run back through [20:29:33] i surprised myself by getting an achievement for bringing the companion cube out of the test chamber. [20:29:44] yeah, i poke around. [20:30:22] i also have rather excellent 3D spatial relationship understanding. so the game is really fun for me that way, the hunting for stuff. [20:30:57] test chamber 18 was easy for me because i used the portal gun to move a hard-light bridge around as a shield. [20:31:06] didn't get shot at once. [20:32:13] Roan: can you review & merge r87450 [20:32:23] ^^ RoanKattouw [20:32:36] !r 87450 [20:32:36] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87450 [20:32:55] bwhaha [20:33:18] I was kind of puzzled why putting it in Hooks didn't fix it :) [20:33:31] but then, it would have been [mwe-upwiz-published] as a not-found message. [20:33:33] Durrr. [20:35:25] Re Portal 2, I haven't played much coop yet [20:35:35] beat single-player last week though [20:36:33] jorm: oh yeah the hard-light shield is fun [20:37:23] neilk_: Done [20:37:27] jorm: I don't know though, I'm finding that almost all the puzzles have only single solutions. I recently tried to go back and see if there were any other ways through and found just a few shortcuts [20:37:35] Back to testing! [20:37:48] Portal 1 did not have that at all [20:38:06] I found that out mostly by playing the challenges or whatever they were called [20:38:19] shut your mouth. [20:38:26] portal 1 had tons of secondary solutions. [20:38:31] That's what I said [20:38:33] which came in useful in the challenges! [20:38:44] I meant Portal 1 didn't have the thing where everything has single solutions [20:38:57] oh, okay. [20:38:59] RoanKattouw: looks good to me. [20:39:09] one of the later ones i got gold on for using only two portals. [20:39:11] Like in that one test chamber, oh look the floor is a deadly swap now, better rethink [20:39:20] *swamp [20:39:26] Nice [20:39:27] deadly swap! [20:39:34] a = b; b =a ; # AAAAAAAHHH [20:39:36] Yeah I found the use-few-portals ones pretty hard [20:39:54] Challenges where they just made the chamber harder, e.g. with the swamp, were the most fun ones [20:40:19] I was a bit disappointed it was all enrichment center stuff again. [20:40:36] but I think it's hard to do portals (technically) with a big map [20:40:46] hence the enclosed-ish nature of each testing area [20:41:12] neilk_: Ready to go to Commons? [20:41:15] yes [20:41:22] Alright [20:42:06] do you guys know the LGTM acronym? [20:42:12] Looks Good To Me? [20:42:15] means "go" :) [20:42:35] we don't do enough cross-reviews here [20:42:50] I say that phrase a lot [20:42:56] I hadn't acronymized it yet [20:43:08] they have t-shirts with LGTM at Google [20:43:34] awjr: PHP Fatal error: Uncaught exception 'ezcMailTransportException' with message 'An error occured while sending or receiving mail. Could not read from the stream. It was probably termina [20:43:36] ted by the host.' in /srv/org.wikimedia.civicrm/sites/all/modules/civicrm/packages/ezc/Mail/src/transports/transport_connection.php:217#012Stack trace:#012#0 /srv/org.wikimedia.civicrm/sites/a [20:43:38] ll/modules/civicrm/packages/ezc/Mail/src/transports/imap/imap_transport.php(2648): ezcMailTransportConnection->getLine()#012#1 /srv/o [20:44:00] ezcMail? That sounds like a skin condition [20:44:04] RoanKattou: thanks, symptom of the upgrade most likely [20:44:11] ah yes i forgot to turn off the cronjobs [20:44:55] !log disabling civicrm and drupal-related cron/hudson jobs [20:44:55] --elephant-- I don't know anything about "log". [20:44:58] damn it [20:44:59] neilk_: eczMail would [20:45:06] !log is Wrong channel! [20:45:06] --elephant-- Successfully added keyword: log [20:46:26] neilk_: Alright, all done [20:46:35] yay [20:47:03] man, this is taking too long [20:47:13] this used to load in a split second when it was all resourceloaderized [20:47:33] You're probably the first to hit it? [20:47:38] still spinning [20:47:52] ok [20:48:26] RoanKattouw: no, now we are getting a few secondary non-resource-ifiable pages and it's slow again :( [20:48:29] WFM [20:48:39] Although mine skips the tutorial [20:48:50] Probably have a cookie for that [20:49:01] How are they non-RL-ifiable? [20:50:17] RoanKattouw: they are wiki pages. Like getting local Commons hackery to do with languages. [20:50:23] Oh, that [20:50:24] and configuration overrides. [20:50:30] OK, now my upload is failing without a clear message [20:50:35] It's probably a dupe of a deleted file [20:50:36] what happened? [20:50:40] no that's fixed [20:50:44] or it's supposed to be [20:50:54] can you send me the file [20:50:54] There is an error symbol [20:51:11] result.error is undefined [20:51:20] strange [20:51:23] Not surprising because... [20:51:26] {"upload":{"result":"Warning","warnings":{"was-deleted":"Testupload.png"},"sessionkey":"trgsdd6y9s49zudy2xqaa6tysfp3orb.png"}} [20:52:44] Alright I'm gonna step up for 30 mins or so [20:52:50] Will check back briefly after that [20:52:53] ok, that is yet another error message, I guess. Other wikis say "duplicate-archive". [20:53:00] *neilk_ shrugs again. [20:53:06] thanks RoanKattouw [20:53:19] you don't even get the "unknown warning" warning? [20:53:26] <^demon> RoanKattouw: Step out, or step up? [20:53:29] RoanKattouw: can you send me that file please? [20:53:47] *step out [20:53:57] neilk_: Nop,e nothing. Will send file and screenshot [20:54:13] oh I see [20:54:20] it's doing result.upload.result = Warning [20:54:33] no wait [20:54:43] it is doing what I expect, why is the code not working. [20:55:06] result.upload.warnings is there... [20:55:09] that makes no sense [20:57:11] neilk_: Done [20:57:17] RoanKattouw neilk_ is the uploadwizard in production been upgraded ? [20:57:25] It was just upgraded [20:57:31] Not 10 minutes ago [20:57:37] :) [20:57:48] GerardM-: somewhat, but I didn't get to push all the changes I wanted [20:57:59] kaldari asked me to ask for more localisations [20:58:11] neilk_ what are the differences people will notice ? [20:58:13] GerardM-: yes, we do need that [20:58:38] GerardM-: the header has more links, like to help out with localizations, known issues, and to go back to the old upload form. [20:58:54] RoanKattouw_away: Can we mark https://bugzilla.wikimedia.org/show_bug.cgi?id=27488 as fixed ? After commiting r85616 you repied there a few times after, so I guess you know it's still open and something needs to be done. What exactly ? [20:58:58] GerardM-: it tries harder to make sure the thumbnails are good, but not hard enough [20:59:05] GerardM-: a few more bugs there [20:59:06] (Implement position: top; for modules) [21:01:09] GerardM-: I owe you a screenshot, though. [21:01:17] GerardM-: we're doing more of an official release next week [21:01:30] hopefully flip it over into default experience [21:04:03] neilk_ I am going to use that as an argument to get the localisations in [21:04:14] thanks :) [21:04:25] spanish, chinese & japanese are particularly lacking. [21:09:33] So, what's up with prototype? [21:13:38] TrevorParscal: Dunno, it's acting very strange [21:14:03] After rating [[AFT Test 3]], [[Special:ArticleFeedback]] turned in to a fatal error swamp. [21:14:09] ( ! ) Fatal error: Call to a member function isKnown() on a non-object in /srv/org/wikimedia/prototype/wikis/rc/includes/Linker.php on line 180 [21:14:21] refresh doesn't fix it [21:14:52] perhaps that should be rc-en (filepath) ? [21:15:07] hm.. alias/rewrite it seems [21:15:31] yeah, it's cause I have to use $wgUser->getSkin()->link() instead of Linker::link() [21:15:52] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87454 [21:16:40] Eyw... right, I didn't test in 1.17wmf1 [21:17:09] crap [21:17:17] that was a bad patch [21:17:20] hld [21:18:20] it only did one of the 3 [21:18:26] and not the one that's actually being used :( [21:19:16] TrevorParscal: So, did you update/livehack on prototype ? It's weird since Special:ArticleFeedback rendered fine before I voted the minute earlier. [21:19:29] yes [21:19:38] but I reverted it to run update [21:19:40] k [21:20:06] ok all better http://prototype.wikimedia.org/release-en/Special:ArticleFeedback [21:21:25] what's the mediawiki number format thing? [21:21:32] $wgLang->something? [21:22:05] yeah [21:22:25] *TrevorParscal looks [21:22:44] ->formatNum [21:22:45] () [21:23:03] function formatNum( $number, $nocommafy = false ) { [21:23:03] #l=2271 Language.php [21:23:25] yeah [21:23:29] found it [21:27:26] /* Protected Static Methods */ [21:27:26] [21:27:26] + protected function formatNumber( $number ) { [21:28:44] and there's no way to use leading zeros i guess... [21:28:46] grrr [21:28:47] http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/languages/Language.php?view=markup#l2260 [21:30:06] Hm .. quick grep suggests wfBaseConvert() [21:30:16] "optionally zero-padding to a minimum column width" [21:30:30] doesnt' appear to be a logical choise as function for zeropadding though [21:47:52] screw $wgLang for now, i'm just using number_format [21:47:59] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87460 [21:48:04] see: http://prototype.wikimedia.org/release-en/Special:ArticleFeedback [21:49:52] design comments, pls. [21:50:07] three face icons: happy, sad, confused: http://elohim.gaijin.com/mbar-icons.png [21:51:04] !svnsearch [21:51:04] --elephant-- I don't know anything about "svnsearch". [21:51:16] !svnsearch is http://toolserver.org/~krinkle/wikimedia-svn-search/ [21:51:16] --elephant-- You don't have permission to do that. [21:54:02] Krinkle: Well it still needs applying of 'position' => 'top' to modules [21:54:17] It's implemented now but not applied [21:54:17] I thought you did that already ? [21:54:40] mediawiki.legacy.diff has position => top [21:55:08] Yeah that's one [21:55:16] I didn't even look at the rest yet [21:55:17] There aren't any other top modules that I know of in core (other than startup and some css) [21:55:31] jorm: what are these for? [21:55:35] Vector/WikiEditor could possibly use some [21:55:51] the MoodBar, one of the -1 to 100 edits projects. [21:55:58] Codereview and LQT for sure RoanKattouw [21:56:13] (php html output needing styling) [21:56:27] FOUC is currently happening to Codereview [21:56:38] white tables changing to coloured tables) [21:56:57] Yes, I thought I'd fixed that [21:57:09] CSS loading is a different issue, distinct from 'top' [21:57:58] Is it ? [21:58:10] Well, 'top' helps [21:58:14] right now it's loaded with themodule, insterted from bottom [21:58:19] jorm: who's mood are you displaying? [21:58:20] But addModuleStyles() is better [21:58:33] the user. they select if they're happy, sad, or confused. [21:58:47] what's the difference between a seperate for 1 module and the general position=top&only=styles ? [21:59:01] (assuming the latter exists) [21:59:32] jorm: http://commons.wikimedia.org/wiki/GNOME_Desktop_icons#Emotes [22:00:00] Yeah. I'm going for more "whimsical"; less "Vulcan" [22:00:11] obviously not all of those go are appropriate [22:00:36] but what i was asking was: "do those express the three moods?" [22:01:03] TrevorParscal: The table on Special:ArticleFeedback, I assume it's not generated live. How do I trigger it (instead of waiting a day) [22:01:20] i can do it for you... [22:01:27] jorm: the confused one is on the edge [22:01:34] not all languages use ? for one [22:01:52] it's pretty hard to screw up happy and sad emoticons [22:02:10] right. i have experimented with wavey lines and raised eyebrows. [22:02:16] Krinkle: The latter doesn't exist [22:02:31] Krinkle: position=top means the styles still get added to the DOM by JS, it just happens earlier [22:02:32] ehn. i'll use these for now for the original document so i don't rathole on this. [22:02:45] jorm: usually confused is displayed using crossed eyes and a puckered offset smile [22:02:59] http://www.iconki.com/icons/Emotions/Yellow-Emotion/YAZOO_SMILIES%20CONFUSED.png [22:03:12] http://3.bp.blogspot.com/_Am_BayAC6XQ/S9umd6WujcI/AAAAAAAAAC4/Z1kT4Qxnom4/s1600/32126-Clipart-Illustration-Of-An-Expressive-Yellow-Smiley-Face-Emoticon-With-One-Big-Eye-Stressed-Out-Or-Nervous.jpg [22:03:14] RoanKattouw: Yeah, but with position=top it'll happen with doc.write, so it should load with or before the page is fully visible. [22:03:21] Yes [22:03:21] and funnily, that's *too* whimsical. [22:03:26] Except that it doesn't load in a noscript env [22:03:30] http://wiki.openttd.org/images/f/fa/Face-Confused-120px.png [22:03:31] Which is bad for things like CR [22:03:33] i have one more similar to the second. [22:03:35] etc. [22:03:43] ...where the styled content is PHP-emitted HTML [22:05:26] if you need styles to be loaded in a non-script environment, you need to use $wgOut->addModuleStyles( 'module name' ); [22:05:27] TrevorParscal: Hey how does adding a UDP backend for ClickTracking sound for a hackathon/train project? [22:05:31] Exactly [22:05:37] RoanKattouw: very good [22:05:59] I'm working on the "don't load until it's visible" part today and tomorrow [22:06:20] OK [22:06:25] it's going to take a lot of testing, because it's pretty important that it works right and it's not an obvious implementation [22:06:30] Also, you mentioned cached responses for anons? [22:06:37] And you're right, the implementation is far from obvious [22:06:40] yeah, I'm going to do that next [22:06:52] How were you gonna do that? smaxage and maxage params? [22:07:40] so, we will say logged in people, or anons with an "i rated!" cookie will use cache=false and not ask for userrating=1 or give any other unique info like user token or whatever [22:07:53] for the initial ArticleFeedbackQuery call [22:08:06] Oh. OK [22:08:09] But that's not enough [22:08:20] ? [22:08:22] You also need to add smaxage and maxage params for the cache=true requests [22:08:28] Assuming cache=true/false is the $.ajax param [22:08:31] ah, yeah that makes sense too [22:08:56] this is why I wanted to drag you into this, so we could make sure we are optimizing as much as possible [22:09:05] I will poke you more when i get to that code [22:09:11] OK [22:09:37] Actually, I guess if you're up for it and jetlag doesn't hit you too hard we could do the UDP thing on Tuesday [22:10:00] what we are trying to avoid is, every page view on all of wikipedia calling ApiQueryArticleFeedback with no caching at all [22:10:04] that would be bad :( [22:10:22] RoanKattouw: sure, we will see how it goes [22:10:52] i also have to sort a couple of bugs out that howie discovered [22:11:05] I guess all design buckets got logged as 1 [22:11:16] Yeah that would suck [22:11:24] The purging thing will be interesting [22:11:45] and for some strange reason the join CTA is shown significantly more often than the others, when it really should be marginally less often since we never show it to logged in users [22:13:40] RoanKattouw: in ApiArticleFeedback.php the bucket param is set to have a min of 1 [22:13:49] Oh, lol [22:13:52] does that mean it would give an error if you passed it a 0? [22:13:52] And you were sending 0? [22:14:00] No, it would probably silently set it to 1 [22:14:00] or just change the 0 to a 1 [22:14:05] well, that makes sense [22:14:09] I'm going to change that min to 0 [22:14:25] Strings will probably all become zero too [22:14:30] (non-numerical stringS) [22:17:05] yup, that was it [22:18:37] one down, 4 to go [22:26:56] robla: Thanks for the note. :-) [22:29:05] Myra is a sockpuppet of Shirley. [22:29:13] *jorm used CheckUser. [22:29:23] Shirley died. [22:29:24] RIP. [22:29:39] Also, abuuuuse. [22:30:10] Hey man. If the president can have his own personal "Team 6" league of assassins, I can abuse CheckUser. [22:30:43] (note that comment does not reflect my personal political views, nor should it be taken as a confession that i used oversight powers. i don't even have them.) [22:31:11] ((Footnote.)) [22:31:32] <^demon|coderevie> I miss Ashlee [22:31:56] i have to listen to Soft Cell because I have them stuck in my head. [22:32:02] don't judge my last.fm feed. [22:32:12] I miss Marybelle sometimes. [22:36:52] Myra: thanks again for weighing in (the note was a "+1" on one of Myra's emails in the parser discussion) [22:39:17] ohlook. another twitter follower who doesn't care about me *at all* and only put me on there because of the celebrity followers and will delete me in 2 days because I don't refollow. [22:39:29] I don't give a shit if you're a tv critic, dude. [22:40:22] alolita, howief: The AFT deployment on Monday is NOT at 11am-noon like that Google Doc of the meeting with CT says, it's at 9am-11am. Noted in the GDoc chat and corrected in the doc. Will also tell CT [22:43:04] *werdna waves [22:43:17] Ohai werdna , how goes [22:43:27] not too bad, just trying to get a festival ticket [22:47:35] is there an alolita about, by the way? [22:48:34] RoanKattouw: interesting -> http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87466 [22:49:34] Already looking [22:49:37] haha, Number.MAX_VALUE [22:49:55] You know [22:50:01] s/round/floor would've gone a long way [22:50:25] In fact, it would probably be fairer [22:50:51] This implementation is potentially infinitesimally unfair because MAX_VALUE may not be a multiple of pitches.length [22:51:34] that's why you use division instead of modulo [22:51:55] http://pastebin.com/gmuP199d [22:52:32] you think I should just modulo list.length? [22:52:47] haha [22:52:48] no [22:52:59] that would have the problems roan was describing :) [22:53:17] I think you should use the original code with round() replaced with floor() [22:53:25] round() penalizes the extrema but floor() does not [22:53:31] res[list[Math.round( Math.random() * Number.MAX_VALUE ) % list.length]]++; [22:53:35] You may or may not have to kill the -1 , I haven't thought that one through [22:53:38] also, % should be replaced with / [22:53:38] ah, i see what you mean [22:54:07] werdna: outside the round? [22:54:33] like Roan says, Number.MAX_VALUE probably isn't a multiple of list.length [22:54:39] No, no division needed [22:54:48] It was using Math.round( Math.random() * ( list.length - 1 ) ) [22:54:56] consider getting a number between 1 and 10, and trying to convert it to one between 1 and 3 [22:54:58] someday, i'm going to figure out how i can NOT be made of meat. [22:55:00] if you do n % 3 [22:55:08] Should be Math.floor( Math.random() * list.length ) ) [22:55:15] RoanKattouw: that makes sense [22:55:21] then 1 will be 1/3 as common as 2 and 3 [22:55:37] Depending on the spec of random() that might return list.length very rarely [22:55:51] werdna: No, that's impossible [22:55:57] You have 10 origin values and 3 target values [22:56:10] A distribution where each target value is equally likely is impossible [22:56:21] that's exactly my point :) [22:56:24] 4-3-3 is the fairest you can do [22:56:33] Oh, sorry [22:56:54] I guess random() theoretically has the same problem [22:57:26] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/87467 [22:57:31] Because in practice there is a finite number of values it can return [22:57:34] But whatever :) [22:58:29] The docs for random() I've found so far are vague [22:58:37] Inconclusive as to whether it can return 1.00000 [22:58:53] well, try it a few billion times and see if it comes out [22:59:20] http://pastebin.com/ruxNnSTF [22:59:36] MDC says it's [0,1) [22:59:38] haha [22:59:45] yeah, I think 1 is not possibe [23:00:04] between 0 and 1 should mean never 0 or 1 [23:00:11] No, it means [0,1) [23:00:15] Possibly 0 but never 1 [23:00:20] Which is what we want here [23:00:37] yeah, that seems right [23:04:06] I think it's designed so that floor( math.random() * arrayindex ) gives you what you need [23:05:57] It is, yes [23:10:14] anybody know if alolita is about? [23:10:27] werdna: hi [23:10:37] howdy [23:17:47] werdna, you're pretty awesome. [23:17:55] I don't know if anyone has told you that lately. [23:18:12] thanks :) [23:18:44] jorm: you're pretty awesome yourself [23:19:26] nah. i'm just a hillbilly out of his pond. [23:22:32] *werdna thinks a bit more than that of jorm [23:23:29] dude, this is IRC, you guys are supposed to be sarcastic and cruel... I think it's part of the freenode AUP. [23:23:38] that's nice of you to think so, but it's not true. [23:24:31] TrevorParscal: it's called being a catalyst, or something [23:24:34] when i was your age, the height of my prowess and ambition was to pretty much get high and play dungeons and dragons in my parent's basement, not be a major contributor to a top-ten property that reshapes the world. [23:24:55] (I did those things successfully.) [23:25:03] I do think jorm is a hillbilly, which is why his stories are so much fun [23:25:06] wait. you're 20 now, right. [23:25:20] The out-of-pond thing is clearly true, I mean he lives in Oakland [23:25:32] so when i was 20, i had a Real Job as a disk jockey. radio and nightclub. [23:25:58] but i still got high and played a lot of D&D. [23:26:07] When I was 18, the job I got to put myself through college was this one :) [23:26:15] \o/ [23:26:19] only it wasn't D&D per-se; it was Warhammer Fantasy Roleplay and Vampire: The Masquerade. [23:26:20] we're living the dream, boys [23:26:23] also some Cyberpunk. [23:26:33] yeah. seriously. [23:26:47] i can't even imagine how cool roan and werdna are going to be in 10 years. [23:26:55] or 5, for that matter. [23:27:28] I'm moving to SF around the end of the year [23:27:36] That should improve my coolness [23:28:00] Roan, you seriously have no idea how cool you are *right now*, in this *very instant*. And that's part of what makes you so fucking cool. [23:28:24] You do need to get your VISA sorted first ;) [23:28:31] Indeed I do [23:28:43] Reedy! Reedy is also pretty damn high on the "cool meter" [23:28:47] What's the story with you, anyway? I heard you're doing the same or at least considering [23:29:03] you guys just need to listen to more Black Sabbath and Led Zeppelin and it'll be hyper-dynamic. [23:29:06] haha [23:29:09] I need to get my medical issues tidied up.. [23:29:41] I'm likely to be in SF for a decent period around the summer, and then look at sorting visas post sept (as apparently after working for a company for a year is easier) [23:30:12] When are you likely to be around in summer? [23:30:29] Not sure exactly. The last uni thing atm is 13th July [23:30:44] ah [23:30:56] I'm looking to head out to SF whenever I can disentangle myself from uni [23:31:00] heh [23:31:00] Probably around the same time [23:31:02] but before that, 4th June I'm just about free at the latest [23:31:04] Then OSCON, then Haifa, then home [23:31:11] so i could get in sometime between those maybe [23:31:39] We should drag you to OSCON but I guess that's expensive [23:31:46] Trevor and I will be going for free [23:32:07] Danese must have some contacts? ;) [23:32:36] oscon's so much more convenient from here :D [23:32:46] "65% off with proof of full-time student status" [23:32:51] Yeah a bit more so [23:33:21] But I would've done OSCON from here no problem [23:33:46] *Reedy shrugs [23:35:05] Damn, sessions only is still more than a grand [23:35:10] $1145 [23:35:23] Even with that discount it's $400 [23:37:06] oreilly ain't cheap. rack up those non-profit and early-bird discounts [23:37:09] and student if you've goty it [23:37:31] They don't allow combining discounts [23:37:43] nooooooooooooo [23:37:50] they must have got wise, i could swear you used to ;) [23:38:05] Nonprofit is 40% off, student is 65% off [23:38:05] i think student is the best discount [23:38:08] yeah [23:38:17] And that's the highest one, yeah [23:38:21] Except 100% for presenters :) [23:38:26] that must be what i was thinking of, mark always got the best discounts before he graduated ;) [23:38:59] Hm looking at the schedule Mon and Tue aren't likely to be interesting [23:39:10] (The big-picture schedule) [23:39:19] no usually those are just the special sessions, and unless something catches you eye skip it [23:39:28] Yeah [23:39:31] one year i did some mobile sessions, but it turned out to suck back then [23:39:44] it was like 'nobody knows what's going on with all these platforms, just kinda wing it' [23:39:54] OSCN Java and OSCON Data have talks on Mon and Tue [23:39:57] Noooo thanks [23:40:10] Also, tutorials, "Node Day", IT leadership [23:41:44] OK, looking over the schedule it looks like there's *lots* of interesting talks [23:41:57] Probably more interesting and relevant stuff than at FOSDEM [23:42:46] awjr: Dude, you're doing an OSCON talk? I had no idea! [23:42:59] the one fosdem i got to was soooooo packed it was insane [23:43:05] could barely get into any talks :D [23:43:08] (I just saw "How to raise millions of dollars using open source software -- Arthur Richards in the corner of my eye) [23:43:17] Oh, we had that for our Wikimedia talk [23:43:21] We got a way-too-small room [23:43:26] And yeah FOSDEM is always very busy [23:43:41] Oh, here we go [23:43:43] Friday at 11:50am [23:43:45] Not a bad timeslot [23:45:53] Ooh, "Git for ages 4 and up" [23:45:59] This is gonna be one interesting conference [23:46:31] haha [23:46:37] Is it too much for 3 year olds? [23:46:45] HOLY SHIT. [23:46:46] I could try on my little sister, she's 4 and a half... [23:46:57] when i was four we only had an atari 400 with 16kb ram, i dont' think it would run git well [23:47:14] Itunes just went from "Tron: Legacy" to Dimmu Borgir's "In Sorte Diablo" [23:47:19] because i had filtered on "legacy" [23:47:32] well, okay. I'll listen to some norwegian speed metal. [23:47:35] heh [23:47:40] RoanKattouw: yeah! im super stoked - you and Trevor are also, right? [23:48:03] Yeah [23:48:05] Anyone else in WMF? [23:48:14] that's it afaik [23:48:34] unfortunately i dont think im going to be able to be in portland for the whole conference :( [23:51:20] haha I love this one [23:51:25] I feel we're kind of in this, or have been [23:51:27] http://www.oscon.com/oscon2011/public/schedule/detail/19140 [23:51:48] oy [23:52:00] haha "- Death sprints: the agile way to approach a death march" [23:52:09] The last sentence is the killer [23:52:17] teehee [23:52:22] It somewhat reminds me of Tim's humor [23:52:43] hahahaha [23:53:22] RoanKattouw, it may be an australian thing :) [23:53:35] Australians are known for dry wit [23:53:43] heya werdna [23:53:53] I was looking to see if she was Australian [23:53:56] they better like dry, i hear their continent is mostly desert [23:53:59] Cause I know them to have great humor [23:54:09] from melbourne according to her web site [23:54:10] But her bio doesn't say [23:54:11] hey awjr [23:54:13] Oh, cool [23:54:18] i remembered she was either english or aussie but wasn't sure :D [23:54:34] I usually remember the difference [23:54:39] Because Australian accents fascinate me :D [23:56:08] we do entire units on it in linguistics [23:56:39] Hmph, the whole personal schedule thing doesn't work for me [23:56:47] I want to take RoanKattouw out bush just so he can stop making fun of *my* acent. [23:56:50] accent* [23:56:55] I don't see any calendar icons or "personal schedule" buttons even when logged in [23:56:58] haha [23:57:13] I don't make fun of it much, do I? [23:57:26] I mean I unintentionally had Liam pretty good once [23:57:36] eh? [23:57:46] He said "car park" and I heard something like "kappa" with a long first a [23:57:57] Or not long, but different [23:58:27] it's something like /k'a-pak/ [23:58:29] Anyway, I asked him to repeat himself and was like OH! Car park! , so he responded mocking my accent with "carrrrrr parrrrrk" [23:58:37] Yeah [23:58:47] It's very weird for a dutch person to have an american accent [23:58:54] most of them have lovely RP accents :p [23:58:54] But I expected something more like /k'a:-pak/ [23:58:59] Indeed they do [23:59:10] And others have horribly-sounding Dutch accents [23:59:17] horrible sounding? [23:59:19] they're brilliant [23:59:21] I can almost always tel [23:59:23] +l [23:59:34] AEvar's girlfriend has a fantastic one [23:59:42] It's fairly typical of a Dutch accent [23:59:49] Not very stron thought [23:59:51] *strong though